CANDIDATE: Dr. Gardner is finishing his research fellowship in infectious diseases at the University of Colorado Health Sciences Center and Denver Public Health. He plans a career as an independent investigator, educator, and clinician in academic infectious diseases. Dr. Gardner's research career development plan includes obtaining an M.S.P.H. degree. He will obtain additional education in behavioral and qualitative research, and questionnaire design. ENVIRONMENT: The research will be conducted at Denver Health, a major care provider for HIV-infected persons in Denver, Colorado. The mentor and co-mentors have extensive experience in epidemiological and clinical research and statistical analysis, have competed successfully for independent funding, and have trained successful researchers and clinicians. BACKGROUND: Antiretroviral adherence has been studied for over 10 years, but generally using a single summary statistic (percent adherence) to summarize adherence behavior. We hypothesize that specific patterns of nonadherence - selective drug taking, frequent missed doses, prolonged periods of time off therapy - can be identified with more detailed study of adherence behavior. A deeper understanding of nonadherence may lead to the development of targeted and durable adherence interventions. The cost of some adherence interventions may be a barrier to their implementation, but the costs associated with poor adherence have not been well studied. SPECIFIC AIMS: 1) to examine pharmacy refill data to acquire a detailed understanding of patterns of nonadherence to antiretrovirals in an unselected clinic population, 2) to apply hypotheses generated through retrospective analysis to prospectively obtained data from clinical trials, 3) to understand the behaviors and system-level barriers which bring about different patterns of pharmacy refills, and 4) to assess the association of antiretroviral adherence with utilization and costs of health care. RELEVANCE: This research aims to understand the barriers that are present within individuals and within the healthcare system that may lead to a failure to take antiretroviral medications as prescribed. It will help clinicians and policy makers understand the association of antiretroviral adherence with the cost of care. If poor adherence to therapy is associated with increased costs of care, clinicians, researchers, and policy makers may be more willing to devote resources to the development of effective adherence interventions.